CAJ | 학술논문

医学 CT 图像的配准一直以来都是医学图形分析与处理中的研究热点之一,对临床诊断具有非常重要的作用.SIFT 算法是一种快速、有效的图像匹配算法,该算法是基于图像的光照、旋转、尺度以及仿射变换不变性,在基于特征点的医学 CT 图像配准中具有较高的实用价值.针对医学 CT 图像配准时匹配特征点数目较少,从而导致估算图像变换参数时精度较低的情况,本文采用结合边缘点集匹配的 SIFT 算法进行医学 CT 图像配准,取得了良好的效果.
의학 CT 도상적배준일직이래도시의학도형분석여처리중적연구열점지일,대림상진단구유비상중요적작용.SIFT 산법시일충쾌속、유효적도상필배산법,해산법시기우도상적광조、선전、척도이급방사변환불변성,재기우특정점적의학 CT 도상배준중구유교고적실용개치.침대의학 CT 도상배준시필배특정점수목교소,종이도치고산도상변환삼수시정도교저적정황,본문채용결합변연점집필배적 SIFT 산법진행의학 CT 도상배준,취득료량호적효과.
Medical CT image registration has long been one of the focus researches in medical graphical analysis and process,Medical CT image registration plays a very important role in the clinical diagnosis. SIFT algorithm is a fast,efficient image matching algorithm which is based on the illumination of the image,rotate,scale,and affine transformation invariance,this algorithm is of high practical value in the medical CT image registration based on feature points. When we conduct medical CT image registration, if the matched feature number is not enough,the calculate accuracy of Image transformation parameters is very low. In this paper,we use a improved SIFT algorithm combination of image edge points to conduct medical CT image registration,and achieved good results.